astuke/HyperTune: Optimization of machine learning hyperparameters in chemical physics

Python code for the optimization of hyperparameters in machine learning (ML), applied to a problem in chemical physics. The ML model at hand is based on kernel ridge regression (KRR) and predicts molecular orbital energies. The hyperparameters in this setup stem from two sources: the KRR method itself and the descriptors for the atomic structure of molecules, resulting in the simultaneous optimization of up two 6 hyperparameters. This repository includes code for three different optimization methods: Bayesian optimization (BO), grid search and random search.